Acoustic Radiation Force Based
Imaging: An Overview
Kathy Nightingale, Michael Wang, Stephen
Rosenzweig, Veronica Rotemberg, Samantha
Lipman, Ned Rouze, Mark Palmeri
Department of Biomedical Engineering
Duke University
Disclosures
Intellectual Property related to radiation
force based imaging technologies
Siemens Medical Solutions, Ultrasound
Division – research agreement providing
equipment and technical support
Learning objectives
• To understand the differences between
acoustic images, qualitative elasticity
images, and quantitative shear wave images
• To understand the tradeoffs between
resolution and accuracy in shear wave
imaging
• To understand the limitations of the
assumptions made by time-of-flight based
algorithms
Elasticity Imaging
Generate images portraying information
about the stiffness (elasticity) of tissue:
1) Mechanical excitation
•
•
•
External
Physiological
Focused acoustic radiation force
2) Image tissue response
•
•
•
Ultrasound
MRI
Optical
3) Generate image of tissue stiffness
•
•
Relative stiffness
Quantify tissue stiffness (shear wave speed
or elastic moduli)
Why image mechanical properties?
• Manual palpation by clinicians – what do they feel?
– Masses (e.g. breast, liver, prostate)
– Pathology (e.g. cirrhotic liver)
– Large inherent mechanical contrast between soft tissues
• Palpation has limitations:
– Physical location
– Size of palpable structure
– Doctor-to-doctor variability (“hard”, “soft”)
– Repeatability
Why use acoustic radiation force?
– Focused within organ of interest
– Small strain
Typical soft tissue material properties
Young’s
Modulus, E
(kPa)
Shear
Modulus, µ
(kPa)
Shear Wave
Speed
(m/s)
Bulk
Modulus, K
(GPa)
Ultrasonic
Wave Speed
(m/s)
Example
~1
~0.3
~0.5
2-2.5
1490-1540
Fat
1-24
0.3-8
0.5-2.8
2-2.5
1490-1540
Liver
3-30
1-10
1-3.2
2-2.5
1490-1540
Skeletal
Muscle
6-45
2-15
1.4-3.9
2-2.5
1490-1540
Prostate
20-150
6.7-50
2.6-7.1
2-2.5
1490-1540
Myocardium
30-300
10-100
3.2-10
2-2.5
1490-1540
Fibrotic Liver
Shear modulus and shear wave speed provide more inherent contrast than
bulk modulus and ultrasonic wave speed.
•Sarvazyan, A.P. (2001). Elastic Properties of Soft Tissue. Handbook of Elastic Properties of Solids, Liquids, and Gases. 3: 107-127
•Skovoroda et al., (1995). Quantitative analysis of the mechanical characteristics of pathologically changed soft biological tissues. Biophysics, 40(6)1359-1364.
•Duck, F.A. (1990). Physical Properties of Tissue, a Comprehensive Reference Book. Academic Press.
•Sandrin et al, (2003) Transient elastography: a new noninvasive method for assessment of hepatic fibrosis. UMB 29(12)1705-1713.
Acoustic Radiation Force
Force generated by a transfer of momentum from an
acoustic wave to the medium through which it is
propagating, caused by absorption (predominantly) and
scattering in soft tissue. Force magnitude typically ~3
g/cm3
F=
2αI ta
c
α=absorption coefficient
Ita = temporal average intensity
c = speed of sound
Pulse 1
Pulse 2
Pulse 3
t
t
t
Nyborg, W. Acoustic Streaming, in Physical Acoustics Vol. IIB, editor: Mason W.P., Academic Press,1965.
FEM: Homogeneous Medium
F=
2αI ta
c
ARFI – look where you push
µ=1 kPa, movie duration = 10 ms
Palmeri et al “A finite element method model of soft tissue response to impulsive acoustic
radiation force”, IEEE UFFC, 52(10): 1699-1712, 2005.
Acoustic Radiation Force Impulse (ARFI)
Imaging (qualitative)
transducer
ARFI
Ultrasound Transducer
B-mode
Displacement inversely proportional to
stiffness
Relative stiffness (as with strain images)
Not operator dependent
Nightingale et al, "ARFI imaging: in vivo demonstration of clinical feasibility", UMB, 28(2): 227-235, 2002
• Radiation force occurs with all wave propagation
• Increased intensity to move microns
• Diagnostic or HIFU transducers
Typical ARFI excitation:
Frequency = 2-6 MHz
Intensity (sppa.5, linear) = 1500 – 3000 W/cm2
Mechanical Index = 1.5-3.0
Duration < 1 msec
Temperature rise = 0.03-0.1 oC
Tissue Displacement = 10-15 µm
ARFI – Prostate Imaging
• Prostate cancer (PCA) facts
– Affects 1/6 men in the US
– 2nd leading cause of cancer death in men
• Prostate cancer diagnosis
– Initially screened through DRE and PSA
– Confirmed through TRUS guided needle biopsy
• PCA not visualized in ultrasound
• Random or systematic sampling
• Low detection rates
• ARFI imaging a potential tool for targeting needle
biopsy and monitoring lesion growth/response to
treatment
http://www.cancer.org/Cancer/ProstateCancer/DetailedGuide/prostate-cancer-key-statistics
Prostate Anatomy and Pathology
Normal Prostate
http://www.ajronline.org/content/188/5/1373/F1.large.jpg
http://www.lab.anhb.uwa.edu.au/mb140/corepages
/malerepro/Images/pro04he.jpg
Adenocarcinoma
Grade 5
Grade 3
http://visualsonline.cancer.gov/preview.cfm?im
ageid=2720&fileformat=jpg
In vivo 3D Prostate ARFI Imaging
Histology (axial plane)
Slice 1
Slice 2
Slice 3
Green – Cancer Gleason 3
Blue – Atrophy
Red – BPH
Black – Urethra
Brown - Verumontanum
Ultrasound
Axial Plane
ARFI w/registered
Pathology
ARFI
B - mode
Coronal Plane
Hsu et al. Proceedings of the IEEE Ultrasonics Symposium, 2011.
ARFI - Monitoring Thermal Ablation
• Thermal ablation increases tissue stiffness
• Ablated tissues no distinct in ultrasound images
• Elasticity methods can monitor thermal
ablation processes:
• Radio Frequency ablation (RFA)
• High intensity focused ultrasound (HIFU)
ablation
• Cardiac ablations are commonly performed
to eliminate aberrant electrical conduction
pathways
In vivo Human Cardiac ARFI imaging of RFA
RF ablation catheter
Position 1
RF ablation catheter
Position 2
•
•
•
•
Human Left Atrium (Roof-line)
AcuNav intra-cardiac transducer and separate RF ablation catheter
AcuNav imaging catheter in fixed position, moved ablation catheter for ARFI imaging
Images courtesy of Dr. Pat Wolf
Shearwave Speed Quantification
• Excite tissue with a dynamic stress:
• Vibrating table or punch1-3
• Acoustic radiation force4, 5-7
• Evaluate resulting tissue response/shear
wave propagation
• Shear wave speed related to shear modulus
(i.e. material stiffness), and structures within
tissue
1Lerner
et. al., 1988; 2Muthupillai et. al., 1995; 3Sandrin et al, UMB, 2003;
et al, UMB, 1998
5Nightingale et al, UMB, 2003; 6Bercoff et al, IEEE UFFC, 2004; 7Chen et al, JASA, 2004
4Sarvazyan
Wave Propagation in Soft Tissues
Ultrasound
(Pressure)
1540 m/s Transverse (Shear ) 1-5 m/s
Particle motion
Wave propagation
Particle motion
Wave propagation
z
x
http://www.kettering.edu/%7Edrussell/Demos/waves/wavemotion.html
Estimate shear wave speed with linear regression
soft
soft
stiff
C=inverse slope
µ=ρc2
stiff
Assumptions:
• Known direction of propagation
• Linear, isotropic, homogeneous
material
Palmeri et al. UMB, 2008.
Liver Biopsy
• Diagnostic gold-standard
– Invasive
• Infection
• Hemorrhage
• Pain
– Limited sampling
– Costly (time and money)
– Not suitable for longitudinal monitoring of
disease progression / resolution
http://www.medandlife.ro/assets/images/Vol%2
0II%20NO%204/generalarticles/fierbinteanu/ima
ge005.jpg
• Can a non-invasive liver stiffness estimate
be used as a surrogate measure of liver
fibrosis?
Shear Modulus vs. Fibrosis Stage
Palmeri et. al., J Hepatology (55), 2011
• 4.24 kPa F0-2:F3-4
threshold
• 90% sensitivity
• 90% specificity
• 0.90 AUC
Commercial Radiation Force Methods
Products now in commercial market (not in US):
– Siemens ‘Virtual Touch Tissue Quantification’
• rEI (qualitative (ARFI) images)
• qEI (quantitative SWS measurements)
• SVI (quantitative images)
• Initial release – abdominal probe, now additional
probes
– Super Sonic Imagine, SSI Aixplorer (quantitative
images)
• Initial release - breast probe, now additional
probes
Liver Stiffness/SWS Quantification/Fibrosis
Over 400 articles in clinical literature evaluating
performance of qEI ™ in the context of liver fibrosis staging
Good diagnostic accuracy for
the noninvasive staging of liver
fibrosis
Siemens – qEI – Local Measures - SWS
Friedrich-Rust, J. Viral Hepatitis, 2012
Toshima, J. Gastroenterol, 2011
Crespo, J. Hepatology, 2012
Sporea, Med. Ultrason, 2010
Heterogeneity in thresholds – why? Sporea et. al., 2010
• Depth within Liver
• Disease etiology (CHC, CHB, NASH/NAFLD)
• Other sources of increases in stiffness (i.e. inflammation,
congestion)
SWS Behavior in Heterogeneous Material
Vertical Layer – resolution and precision
∆ RMS (m/s)
regression kernel size:
5 mm kernel
Resolution (mm)
2 mm kernel
Matched C-plane In Vivo Prostate Images
ARFI (Qualitative)
SWS (Quantitative 0-4 m/s)
• Quantitative SWS image is lower resolution
• Concordance between darker ARFI regions and higher SWSs
SSI – Multi-center Breast Lesion Evaluation
Breast Fibroadenoma
BI-RADS 5
BI-RADS 4a
Ductal Carcinomas
BI-RADS 3
939 breast masses; limited SSI to
evaluation of BIRADS 3 and 4a:
• Increased specificity of breast mass
assessment from 61.1% (397 of 650)
to 78.5% (510 of 650), with P<.001
• Insignificant improvement in sensitivity
Berg et. al, Radiology: 262(2); 2012
Summary – Radiation Force Based
Elasticity Imaging
• Clinically available
– Qualitative methods (ARFI imaging)
– Quantitative methods (shear wave speed)
• Need large-scale clinical studies and research
validation of the quantitative methods
– monitoring disease progression?
– monitoring response to therapy?
• Standardization among manufacturers –
RSNA/QIBA efforts
Acknowledgements
• NIH NIBIB R01EB002132
• NIH NCI R01CA142824
• Siemens Medical Solutions, USA, Inc.,
Ultrasound Division
Duke ARFI/Ultrasound Team
3D Shear Wave Imaging Setup
HIFU Transducer
4Z1C
3D
probe
ARFI
push
muscle
fibers
Shear Wave Propagation in Excised Canine Muscle
Muscle SWS (m/s)
ct H = 3.9 m/s
ct ⊥ = 2.5 m/s
Matched C-plane In Vivo Prostate Images
ARFI (Qualitative)
SWS (Quantitative 0-6 m/s)
• Quantitative SWS image is lower resolution
• Concordance between dark ARFI regions and higher SWS